Counting Words

Inspired by Colin's digging in the Web Search Platform we dug some more. This time we just counted all words found on the web. Some interesting things came up. The 5th most popular 10-letter word on the Web? California. The 7th most popular 12-letter word? Pennsylvania The most popular 15-letter word? Zaregistrirovan.

Zaregistrirovan? Yes. The 15 letter list is kind of funny. For your amusement: zaregistrirovan. partnerprogramm. netutodeanatano. anmeldungsdatum. zuckerkrankheit. recommendations. kontaktanzeigen. nenutzergruppen, mitgliederliste, veranstaltungen.

The graph below represents the most popular 5-letter words. Judging by this list of words forum posts must make up a huge chunk of the Web.

forum - 2,238,602,797
about - 1,351,249,584
other - 962,689,569
posts - 946,766,048
video - 839,711,215
email - 775,440,102
found - 751,607,505
phpbb - 741,762,343
board - 723,989,141
click - 700,040,348


The next graph is the most popular 6-letter words. The Web must be getting more civilized; somehow the word "please" made into the top 3:

search - 1,728,203,440
online - 1,665,987,136
please - 586,515,321
posted - 561,419,032
people - 540,502,379
cannot - 501,884,374
rights - 499,446,153
topics - 478,526,379
should - 474,777,139
server - 469,748,101


The last graph here is the most popular 7-letter words:

contact - 866,133,410
service - 715,363,809
powered - 587,313,121
general - 463,853,406
private - 440,264,813
company - 436,092,563
website - 432,281,826
message - 422,914,185
product - 406,084,827
support - 387,550,901



It would be interesting to see how this list has changed over the years. I'd bet my last nickel that the most popular 7-letter words in 1996 weren't Contact Service.

Comments - Permalink

Coordinates, Round 2

We had some curious people ask what was buried in that mass of red dots in the lower left corner of the graph (see previous post), so we zoomed in on it. Another surprise! There are some geographic coordinates buried in there. So, here it is, a zoom of lower left corner of the graph, showing a somewhat familiar looking outline. If you look closely you'll even see the path of Katrina in there.

We still don't know exactly what each data point represents. It would be fun to put up another graph and make each data point clickable with a link to its location on the web.

Comments

Fun with the Web Search Platform

Our program manager Colin Saunders was playing with the Web Search Platform over the weekend. He didn't have a particularly inspiring idea, so he did the first thing that came to his head and plotted the results on a graph. The results are, ummm... interesting.

His idea? Mine for any set of numbers that looked like coordinates. If it was a set of numbers, inside parentheses, separated by commas, like (1024,768) he stored it. He ran the job across approximately 10% of the Web (for about $200) placed it into a table, then put the coordinates on the graph you see here.

I'll leave the comments to you...

Comments